Subject: European tripoint lat & long list
Date: May 25, 2002 @ 22:19
Author: Grant Hutchison ("Grant Hutchison" <granthutchison@...>)
Prev    Post in Topic    Next [All Posts]
Prev    Post in Time    Next


I've recently been downloading data from

http://rimmer.ngdc.noaa.gov/coast/getcoast.html

and extracting and collating international boundary data. This is tricky,
since the boundary data are stored (at least in the downloaded text files)
as numerous short segments, up to sixty per border segement. Not all are
orientated in the same sense, so it takes a bit of effort to pick out all
the data for a particular boundary and to collate it into a single sequence
of longitude and latitude coordinates. I've also trimmed out multiple
duplicated datapoints.

I've used the World Vector Shoreline (WVS) data for longstanding borders,
and have assembled borders in the former Yugoslavia and USSR from 1970s
internal boundary data in the World Data Bank II (WDB). There are bound to
be discrepancies between the real world and these old data, but I've
uploaded a collated file for European borders (Europe.txt) to the files
area - the more data the better, in my opinion, provided we are aware of the
limitations, and this sort of information doesn't seem to be readily
available from any other source.

The file consists of a sequence of headers identifying each border segement
(assembled from ISO digraphs in the usual way), each followed by a long list
of longitudes and latitudes. When border segments come in two parts
separated by another land boundary, they are differentiated by "n" and "s"
or "w" and "e" suffixes in the header. When the land portions of a single
boundary are interrupted by an arm of the sea (denl, depl, bahr) border
segments are suffixed "1" & "2".

Although the lats and longs are provided to a spurious six decimal places
(10cm!), I think they're probably only accurate to a kilometre or so - this
estimate is based on the fact that the line of international borders in WVS
and in WDB often differ to this extent or more. This causes mismatches in
tripoints that are defined by a mixture of WDB and WVS data. I've left this
unresolved, simply breaking the WVS border at the closest point to the
termination of a WDB border. For instance, I've broken the old
Czechoslovakian/Austrian border (derived from WVS) into atcz and atsk
segments - the position of the break was determined by the southern
termination of czsk (derived from WDB).

Tripoint data are easily derivable. Below, I give what I think is a complete
list of 44 European tripoints, together with their longitude and latitude
from WVS/WDB. I make no defence of their accuracy - I just report what the
database says, and hope it may be of use or interest.



Grant

=================================

Tripoint Longitude Latitude

adesfre 1.734605 42.502944

adesfrw 1.443304 42.600338

algrmk 20.991854 40.858106

almkyu 20.585558 41.879713

atchde 9.559833 47.539252

atchit 10.482139 46.859257

atchlin 9.541352 47.264673

atchlis 9.620557 47.058738

atczde 13.843980 48.776035

atczsk 16.947743 48.621657

athusi 16.116010 46.872458

athusk 17.161232 48.008913

atitsi 13.719305 46.519846

bahryun 19.039727 44.861367

bahryus 18.455585 42.565795

bedelu 6.133160 50.128398

bedenl 6.024326 50.752950

befrlu 5.834232 49.540516

bggrmk 22.944127 41.335687

bggrtr 26.362292 41.706194

bgmkyu 22.363579 42.320038

bgroyu 22.672481 44.212027

byltlv 26.613330 55.674703

byltpl 23.504141 53.946919

bylvru 28.168035 56.150304

byplua 23.604688 51.527772

byruua 31.783922 52.108027

chdefr 7.590838 47.582669

chfrit 7.046372 45.929910

czdepl 14.831118 50.834502

czplsk 18.851393 49.517488

defrlu 6.368137 49.461310

eelvru 27.371945 57.335523

finoru 28.912129 69.054744

finose 20.562383 69.063544

hrhusi 16.607745 46.476136

hrhuyu 18.904710 45.937244

huroua 22.894550 47.951122

huroyu 20.271082 46.125871

huskua 22.159404 48.398488

ltplru 22.770241 54.383870

mdrouan 26.635625 48.266918

mdrouas 28.217759 45.468317

plskua 22.570100 49.087870